of brain dysfunction, beginning with Phineas Gage in 1848 and Paul Broca's observations of Tan in. 1861 to the "split-brain" callosotomy patients first described ...
Journal of Neurotherapy Volume 13. Number 4, 2009
CONTENTS EDITORIAL Psychological EEG Analysis David A. Kaiser, PhD
193
SCIENTIFIC ARTICLES Neuromodulator}- Approaches for Chronic Pain Management: Research Findings and Clinical Implications Mark P. Jensen, PhD. Leslie //. Sherlin, PhD. Shahin Hakimian, MD. and Felipe Fregni, MD. PhD
196
Alpha Neurofeedback Training for Performance Enhancement: Reviewing the Methodology D. Vernon, PhD. T. Dempster. BSc. O. Bazanova, PhD, /V. Rutterford. PhD. M. Pasqualini. PhD. and S. Andersen. PhD
214
PERSPECTIVES Rhythms of Healing: A Case Study Richard L. Jacobs, PsyD
228
PROCEEDINGS OF THE 2009 ISNR CONFERENCE Selected Abstracts of Conference Presentations at the 2009 International Society lor Neurofeedback and Research (ISNR) 17th Annual Conference, Indianapolis, Indiana 239 PROCEEDINGS OF THE 2009 SABA CONFERENCE Abstracts of Conference Presentations at the 2009 Society for the Advancement of Brain Analysis (SABA) 8th Annual Conference. Marineland, Saint Augustine, Florida 277
Following Page: 283 Title Page for Volume 13 Author Index for Volume 13 Journal of Neurotherapy, 13:193-195, 2009 Copyright 0 Taylor & Francis Group. LLC ISSN: 1087-4208 print/1530-017X online 10.1080/10874200903338570
ïi DontloHno B< ^ DOl:
?
EDITORIAL
Psychological EEG Analysis Recently a colleague of mine used quantitative EEG to help him determine whether a man convicted of felony murder deserved the death penalty. The question was whether this young man had sufficient mental capability to control himself. My friend, a psychologist, used a number of psychological instruments, but the QEEG results were by far the clearest, revealing significant functional immaturity missed by conventional neuropsych testing. Anyone trained in psychological EEG analysis would come to the same conclusion as my colleague, that the convict's frontal lobes were significantly functionally undercon- nected. in nearly every index of EEG connectivity, that his brain activity resembled that of a teenager more than that of an adult in his early 30s. I was Hown in to testify, to support the application of EEG analysis for this purpose. The practice of law is the practice of science, an ideal slowly actualized through iteration. In both realms progress is slowed by ambition and error, but such is to be expected of any human endeavor. It is also clear to me that law must incorporate scientific understanding of human behavior if it is to survive, and functional neuroimaging in particular poses a unique challenge to its existence. Law and science both rely on prccedcnts. though wc have an advantage over law in that we are able to repeat an experiment as many times as necessary to understand human behavior and they must judge a single instance. We can replicate an action and change participants, setting, lighting, anything of interest, but law is given a single outcome and force to play the cards it was dealt. In my testimony I explained to the judge and trial lawyers how individuals trained in Psychology interpret neuroelectromagnetism differently from those trained in Neurology, that EF.G is a tool
shared by both fields and one group may use it for X and another for Y and either may be valid in their use and interpretation. Of course I didn't use algebra to make my point but analogies, as a verbal reasoner. the judge, was the audience I hoped to convince. EEG is a tool, and like any tool it may be used in different ways by different people. A judge uses a hammer to maintain order and a carpenter uses it to build a house, but we wouldn't force all hammers to be round at both ends, or require a metal claw to protrude from a gavel, would we? Hammers come in all shapes and sizes, and like EEG they can be wielded in different ways. Same tool, different purposes. The opposing view, expressed by the DA, was the neurologist's view of this tool, that EEG can be used only to identify structural problems in the brain, a stance so misguided and limited that it should evoke pity in the reader. The study of neuroelectromagnetism is closer to cosmology than medicine in the way we are doing it. We are investigating how human thought creates light, currents of electromagnetism. the generation of orderly energy, and how that regulated energy controls and manipulates matter, the body. Ln fact Hans Berger created the first EEG equipment in the 1920s to study thought processes (mostly in his son) and
194
JOURNAL OF N FU ROTH ERA r Y only later did he adapt his technology to study late 1960s, and all serious sciences that employed epilepsy and other brain disorders. frequency analysis seismology, acoustics, physical The defense team phoned me a few days after my oceanography responded to this change... except testimony and asked me to help undermine the neurology, apparently. credibility of the opposing expert, a well-known During most of my examination and crossneurologist and friend of a friend. 1 told them to ask examination the computer screen behind me showed the witness questions relevant to psychological a Vend diagram of Psychology and Neurology, the investigations of EEG. For instance, what would we study of mental functions and behavior and the study expect to see in an F.FG record of an individual of disorders of the nervous system, respectively, during a lexical decision task, with or without with their intersection being "EEG" (see Figure 1). lateralized presentations... that's easy enough. And The assistant DA asked me time and again about the what happens at posterior electrode sites when a use of the EEG in diagnosing organic disorders, and gambler playing blackjack considers himself ahead time and again I responded that we were using EEG or beating the house? Most of my undergraduates to study mental operations and abilities. My can answer these questions, and these are just two of testimony was being evaluated under the Daubcrt the many psychological applications associated with decision (1994). which produced guidelines for EEC analysis. More than 42,000 quantitative EEG evaluating scientific evidence and testimony in court papers have been published since 1965. and a survey trials. A technique or theory must be accepted by the of last year's publications revealed three times more relevant scientific community and governed by EEG research on issues under the umbrella of explicit rules to be viewed as credible evidence in a Psychology—attention, sleep, unconsciousness. court of law. Aware of this decision, I explained how animal behavior—than under the smaller parasol of science, like law, is not homogeneous or uniform but Neurology. In other words, identifying and studying an assemblage of independent disparate groups, each organic disorders with EEG has been a minority its own school of thought, largely inert to the application of this technology for my entire lifetime. successes and failure of adjacent fields. Each science The DA asked me whether I knew the neurology makes its own rules, and only rarely do mies of one standard for sampling rates was 200 Hz. and I field extend or cascade into another. Unlike law. laughed when he told me the number, disbelief to the science has remarkably few rules shared across point of mockery. Why would any BEG science rely disciplines, but they are the following: a on a base-10 number? I asked aloud. The defense measurement must be repeatable. a theory loved my response and I continued to explain to the falsifiable, an inference logical, all tools are unhappy DA how such a standard was archaic, 40 considered imperfect, and communication between years out of date, a relic from the hey-day of Grass practitioners or to the public must be honest and and Gibbs in the late 1930s, perhaps, or established transparent. Those are our by Molly Brazier and her crew in the 1950s. A base-10 sampling rate was like using horse-andbuggy rules to control highway traffic. The Fast Fourier transform (FFT) was invented in 1965. and no serious scientific group would rely on a base-10 value after its occurrence. as it would sacrifice accuracy, speed, and communication, the trifecta of scientific investigation. We have been the power ol 2 since 1965. with rates set to 128, 256. 512 samples per second for a reason. The FFT is a clever and highly efficient algorithm for quantifying frequency information within a time scries, light years ahead of any spectral estimation technique in vogue prior to 1965. The discrete Fourier transform (DFT), available since the 19th century, for instance, requires endless iterations of trigonometry and floating-point operations, a dicey and memory-intensive operation 40 years ago. whereas the FFT is an NlogN algorithm, which is geek speak for "Hello, gorgeous!" It was superior in every way, a major algorithmic breakthrough when it hit the world in the
FIGURE 1. Use of EEG technology shared by disciplines.
Editorial
Mental Organic correspondence correspondence with behavior with disease standards, by the way. Unfortunately the other day I noticed one of our "practitioners" breaking Rule 4. using the phrase "zero error* on his Web site to describe his analysis. We should remind ourselves that making a claim of zero error is not science but propaganda. My take-home message from this experience was that the International Society for Neurofeedback and Research (1SNR) needs its own standards, which it is working on, and the recognition that neurofeedback instruments arc likely misclassified by the Food and Drug Administration as neurological therapeutic devices ($ 882.5050). The Food and Drug Administration does not regulate psychological tools such as the Minnesota Multiphasic Personality Inventory, Beck Depression Inventory, or l est of Variables of Attention, nor does it classify video games as food or drug, so how does computer-interface technology that encourages mental exercise and psychological change fall under Journal of Neurol herapy, 13:196-213. 2009 Copyright ft Taylor & Francis Group. LLC ISSN: 1087-4208 print/1530-017X online LX)I: 10.1080/10874200903334371
the aegis of a government agency dedicated to drug safety? And if it does, why aren't World of Warcraft and other addictive video games similarly regulated? 1 think ISNR needs to establish its own system of evaluating the claims, safety, and efficacy of neurofeedback equipment. In this issue of the journal David Vernon and colleagues weigh in on the general195 principles of neurotherapy. and Mark Jensen and his colleagues discuss opportunities for neurotherapy in pain management. Also included in this issue are the proceedings from our recent conference in Indiana, which was like a Roman marketplace of new ideas and new technologies in the service of mental health. Along with ISNR's proceedings, abstracts from the Society for the Advancement of Brain Analysis (SABA), a daughter group of ISNR. are also included. Finally the Perspectives section reappears, a section for clinical narratives. Storytelling is an important part of scientific investigation as stories arc often the best teachers. There are many who prefer examples to abstract principles, and in fact behavioral neurolopr owes its existence to case studies, that is. singular examples of brain dysfunction, beginning with Phineas Gage in 1848 and Paul Broca's observations of Tan in 1861 to the "split-brain" callosotomy patients first described at Caltech in 1962 to the uniquely brain-injured patients known only by initials, such as the late H.M. We hope the inclusion of clinical narratives in this journal provides further insights into the use of neurotherapy. David A. Kaiser, PhD Senior Editor , ' 9e r^E.Fn^Ccoop ß
t
SCIENTIFIC ARTICLES
Neuromodulatory Approaches for Chronic Pain Management: Research Findings and Clinical Implications Mark P. Jensen, PhD Leslie H. Sherlin, PhD Shahin Hakimian. MD Felipe Fregni, MD, PhD
ABSTRACT. Two lines of evidence provide preliminary support for the role thai brain state, measured via electroencephalogram (EEG). may play in chronic pain. First, research has identified a link between brain EEG activity and the experience of pain. Second, there arc a number of published studies documenting the beneficial effects of interventions that impact the cortical activity associated with chronic pain. These interventions include neuro- bchavioral treatments such as neurofeedback and hypnosis as well as invasive and non-invasive brain stimulation. Preliminary data showing the efficacy of neuromodulator^- strategies for treating pain provides compelling reason to examine how conical activity (as measured by BEG) may underlie the experience of pain. Existing data already suggest specific approaches that neurofeedback clinicians might consider when treating patients with chronic pain. Reciprocally, observations by neurofeedback practitioners could provide important ease data that could foster the design of more definitive randomized clinical trials using such strategies lor the treatment of chronic pain.
Murk. P Jensen is affiliated with the Department of Rehabilitation Medicine, University of Washington School of Medicine. I .eslic 11. Sherlin is affiliated with '.lie Department of Psychology, University of Phoenix; Southwest College of Naturopathic Mcdicine & Health Sciences; Northern Arizona University; and Nova Tech EEG. Inc. Shahin Hakimian is affiliated with the Department of Neurology, University ofWashington School of Medicine. Felipe Fregni is affiliated with the Center for Noninvasive Brain Stimulation. Harvard Medical School. Some of the material in this article was presented in a previous article and we acknowledge its original publication (Jensen. Hakimian. Sherlin. & Fregni, 2008). Address correspondence :o: Mark P. Jensen. PhD. Department ol Rehabilitation Medicine. Box 356490. University ofWashington School of Medicine. Seattle. WA 98195-6490 (E-mail: mjcnsen.gu.washington cdu).
KEYWORDS. Chronic pain, hypnosis, neurofeedback, neuromodulation. tDCS, transcranial direct current stimulation
INTRODUCTION AND OVERVIEW
Chronic pain is now known to produce plastic changes in an extensive neural network that involves
areas associated with somatosensory and emotional-affective processing. An understanding of the specific neurophy- siological changes that are associated with chronic pain could contribute Scientific Articlesto the development of novel treatments aimed at central nervous system (CNS) modulation. Research from a number of sources suggests a link between electroencephalographic (EEG) activity and the experience of pain. As described in more detail in the section that follows, this research suggests that (a) in most otherwise healthy individuals, the subjective experience of pain is associated with relatively lower amplitudes of alpha activity and relatively higher amplitudes of beta activity and (b) individuals with chronic pain associated with neurological disorders, such as spinal cord injury, evidence higher amplitudes of theta activity and lower amplitudes of alpha activity than individuals who do not have chronic pain. The purpose of this article is to summarize the extant evidence concerning the associations between EEG-asscssed brain activity and pain, and then to discuss the implications of these findings for understanding the mechanisms of treatments that modulate cortical activity such as neurofeedback. hypnosis, and noninvasive brain stimulation. We also include a discussion of neurofeedback treatment approaches that may be used for chronic pain management that are based, in part, on the available evidence concerning the EEG activity most closely associated with pain. The goals of this article are therefore to (a) alert clinicians to recent findings on this field that might contribute to the development of effective chronic pain treatments now and in the future and (b) encourage more research to examine the efficacy of neuromodulator interventions and to understand the mechanisms of their effects. INTEGRA T/ON AND PLASTICITY IN THE NEUROPHYSIOLOGY OF PAIN Before the middle of the 20th century, pain was viewed primarily as a simple reflexive response to physical damage. In that view, pain information (nociception) was thought to be transmitted directly from nerves in damaged tissue through a single channel directly to a "pain center" in the brain. When seeking to diagnosis or understand pain from this model, the area of primary interest was the periphery, that is, where pain was thought to originate from; the notion was that i;real pain" was mostly related to the amount of physical damage or inflammation that occurred in peripheral structures. The brain was viewed as an essentially passive recipient of sensory information. An initial important turning point in our understanding of pain occurred with the publication of the gale control theory of pain (Melzaek & Wall. 1965). This theory provided a model for how
nociceptive input is influenced and modulated in the spinal cord before it reaches the cortex and then, ultimately, leads to the experience of pain. In addition, pain associated with neurological lesions 197 such as spinal cord injury and stroke started to be viewed in the context of this new understanding that the brain plays a major role in chronic pain. More recently, a significant increase in the use of advanced neuroimaging methods for understanding the brain mechanisms involved in pain has helped to confirm this notion that chronic pain can result from dysfunction in the central nervous structures. As a result of recent research, pain is now understood as an experience that is influenced by a dynamic series of multiple interlocking neurophysiological mechanisms that modulate nociceptive information at many levels, including supraspinal sites such as the cortex (Apkarian. Bushnell, Trecde. & Zubieta, 2005; Craig. 2003a, b; DeI.eo, 2006; Katz & Rothenberg, 2005: Melzaek, Coderre. Katz, & Vaccarino, 2001; Miltner & Weiss, 1998; Tina7zi et al.f 2000).
JOURNAL OF NEUROTHE R A P Y Moreover, we now know that, in addition to the effects caused by input from the periphery (e.g.. nociceptive information that is sent along the A-dclta, A-beta. and C fibers to the spinal cord and into the CNS) supraspinal structures such as the primary and secondary somatosensory cortex., insular cortex, anterior cingulate cortex, prefrontal cortex, and thalamic nuclei all work together to represent and modulate the experience of pain (see Figure 1). Thus, although activity in the peripheral nervous system and the spinal cord certainly can play an important role in the experience of pain, the key role that multiple conical and subcortical sites play in the perception and emotional response to pain is now more clearly acknowledged and understood. Research has also shown that nociceptive input, FIGURE 1. The primary nervous system structures involved in the processing and experience of pain. Reprinted with permission.
via neural adaptation, may modify neural networks' -To Anterior circulate cortex
responses to repeated stimuli (Flor. 2003; Katz & Melzack. 1990). For example, sensitivity to noxious
198 stimulation increases as a result of ongoing nociceptive input (Bromm & Lorenz, 1998); in other words, the experience of pain itself makes the
Sp-nottialnmic tract
CNS more sensitive to pain—this phenomenon is called nervous system sensitization. This hypersensitivity may have an evolutionary advantage as increased pain can promoteArticles healing by Scientific inducing behavioral changes, compelling a person to take special care of injured anatomy. However, this mechanism can at the same time extend discomfort and suffering beyond the time it takes for healing to occur, and potentially contribute to the development of a chronic condition. Some studies have identified the CNS changes that occur with chronic pain as a potential target of treatment to produce pain relief; interventions that reprogram or interrupt central sensitization, at the cortical level, could also possibly provide significant relief for some individuals with chronic pain (Flor, Braun. F.lbert, & Birbaumer, 1997; Maihofner. Handwerker, Neundorfer, & Birklein, 2003; Pleger et al., 2004; Tinazzi et al., 2000). Another important issue related to focusing on supraspinal systems as a target for chronic pain interventions is that the brain should be viewed as a two-way system, in which the processing of information results in changes in efferent systems ihrough top down mechanisms such as regulation of endocrine and immune systems. It is therefore also possible that modulation of pain by altering CNS activity directly may promote health by stimulating salutogenic mechanisms (Fregni, Pascual-Iveone. & Freedman. 2007). MEASURING THE NEUROPH YSrOE OGICA L CORRELA TES OF PAIN Researchers have used a number of tools for studying and identifying the neurophysio- logical correlates of pain. Functional magnetic resonance imaging (fMRI) has been used to measure localized changes in blood flow in the brain (and therefore neuronal activity) associated with pain. Positron emission tomography (PET) has been used to assess localized brain metabolic changes induced by aversive stimuli. EEGs have also been used to infer changes in brain electrical field activity after experience of pain. Finally, transcranial magnetic stimulation ( I MS) has been used in two different manners to assess chronic pain: (a) via single and paired pulse TMS to assess cortical excitability changes associated with chronic pain (e.g.. Lefaucher. Drouot. Menard-Lefaucheur. Keravel, & Nguyen. 2006) and (b) via repetitive TMS (rTMS) to functionally and transiently disrupt activity in the anatomical sites of pain experience (inducing "virtual lesions" that can link brain anatomy and its functional behavior) or by facilitating activity ¿titer
the end of stimulation. Each approach has its advantages and limitations. A primary strength of fMRI and PET is that these imaging strategies can localize activity 199 throughout the brain. For fMRI. localization can occur at a relatively high degree of spatial resolution. However, as measures of the correlates of experience, the ability of these imaging methods to establish cause-effect relationships is limited. Also, with these methods of neuroimaging. temporal resolution is relatively poor. Measurement of cortical excitability via single and paired pulse TMS has the advantage of providing reliable functional measurements. However, this approach is limited to motor (and to a less extent, visual) cor- tcx. rTMS. on the other hand, has the advantage of having a good temporal resolution, and it can also allow for inferences about causal relationships. Because pain is a complex experience associated with activation in an extensive neural network, some of the temporal resolution of rTMS may be lost. In addition, this method assesses function by instituting changes; it does not itself directly measure activity (Pascual-Leone et al.. 1998) except when using single and paired pulse TMS. EEG measures, although less commonly employed than fMRI and PET studies, can provide information complementary to fMRI. PET. and TMS/rTMS. Specifically, EEG can assess cortical rhythms in specific frequency bands, which are associated with different brain states. Significant support for the potential of EEG measures for contributing to our understanding of pain processing comes from evidence that the power of different EEG band widths has been shown to be associated with pain severity. Data from acute (induced) pain models to study the
effects of pain on HEG measures have shown a patients' EEG patterns, including the differences in consistent pattern. Specifically, these data have theta bands, normalized after 12 months. The 200 JOl RNAL OF NEUROTHERA P Y shown that with more intense painful stimulation, all authors of this study hypothesized that the EEG EEG frequencies increase in power, but beta differences found may have been the result of a frequencies increase relatively more than other thalamocortical effects (i.e.. postulated increased bandwidths. and the relative power of alpha activity thalamus theta activity resulting from decreased tends to decrease (Bromm. Ganzel, Herrmann. input into the thalamus. which then contributes to an Neier, & Scharein. 1986: Bromm. Meier. & increase in theta activity throughout the cortex). Scharein, 1986; Chang. Arendt-Nielsen. These findings were replicated in a second Graven-Nielsen. Svenson. & Chen, 2001: Chen. D sample of patients with chronic neuropathic pain work in, & Drangsholt. 1983: Huber. Bartling, who were also candidates for a CLT (Stem, Pachur, Woikowsky-Biedau, & Lautenbacher. Jeanmonod. & Sarnthein. 2006). In this second 2006: see also reviews by Bromm & Lorenz. 1998; study, the patients with neuropathic pain were found Chen, 1993, 2001): in summary, more acute pain to have higher levels of both theta (6-9Hz) and beta appears to lead to more relative beta and less relative il2-16Hz) activity relative to healthy controls. Also, alpha activity. At the same time, this research indi- and consistent with the findings of Sarnthein ct al. cates that acute pain relief is associated with (2006), successful treatment with a CLT resulted in decreases in the relative power of beta activity and gradual decreases in EEG activity in the theta and increases in the relative power of alpha wave beta range over the course of 12 months. The authors activity (sec also Kakigi el al.. 2005; Pclletier & concluded that the "spontaneous, ongoing, Pepcr, 1977). frequency-specific over-activations may... serve as There is much less research examining the effects an anatomo-physiological hallmark of the processes of chronic pain on EEG measures. The findings from underlying chronic neurogenic pain" (Stern et al.. the few studies that have examined EEG activity in 2006. p. 721). patients with chronic pain are generally consistent A more recent study compared EEG- assessed with those from acute (induced) pain studies, with cortical activity in three groups: (a) patients with one interesting exception: A significant increase in spinal cord injury and chronic pain (/? — 8), (b) relative very slow (theta) activity is found in patients with spinal cord injury without pain (w = 8), individuals with chronic pain. One of the first of and (c) healthy controls (n = 16: Boord et al., 2008). these studies compared resting EEG bandwidth Consistent with the finding from both Stern et al. activity between 15 patients presenting with a (2006) and Sarnthein et al. (2006), these variety of chronic neuropathic pain problems (who investigators found that the peak activity in the were also candidates for a central lateral chronic pain sample was in the theta range, on thalamotomy (C'LTJ) with 15 otherwise healthy average, whereas peak activity in the nonpain individuals (Sarnthein. Stem, Aufenberg, Rousson, samples was in the alpha range. Similar to the & Jeanmonod, 2006). About half of the patient findings of Sarnthein et al., Boord et al. reported that group was taking centrally acting medications (e.g., the use of centrally sedatives, opioids, antiepileptics, and antidepressants). Consistent with the acute pain research, and at rest, the patients with chronic pain had elevations in all EEG frequency bandwidths. Moreover, and also consistent with the acute pain research, patients with pain had more relative (relative to overall activity) beta activity and less relative alpha activity. However, unlike the findings from acute pain research, the patients with chronic neuropathic pain in this study also evidenced more absolute and relative slower activity (in theta band). The patients with pain who were taking centrally acting medications showed the same pattern of findings as those with pain who were not taking any medications, although the differences between these (medication-taking) patients and the otherwise healthy participants were less pronounced than those between the controls and patients with pain who were not taking medications. Of interest, following the C-'LT, which resulted in pain decreases, the
acting medications among the patients with pain was associated with some differences in EEG activity, relative to those who were not taking medications, such that EEG activity in those taking Articles medications Scientific was shifted a little in the direction of activity that was a little more like those without pain. However, this effect occurred at only 3 (P3, P7. and Pz) of 14 assessment sites. Also, overall, the differences observed between those with and without pain were found over many sites, suggesting a diffuse effect of pain and raising the question of whether some cortical areas may contribute more or less to the differences found than others. NEUROFEEDBACK TRAINING AND PAIN RELIEF Based on the evidence showing that EEG activity is linked to the experience of pain, it is possible that neurofeedback training could be used to Leach patients to increase or decrease relative power of different EEG band widths as a way to treat chronic pain, specifically to alter brain activity in such a way that it reflects the EEG activity that has been shown to be associated with less pain (Batty. Bonnington, Tang. Hawken. & Gruzclier, 2006; Egner. Strawson, & Gruzclier. 2002; Vernon et al., 2003). Preliminary evidence is consistent with this hypothesis (see Table 1). One case study, one laboratory study, and a case series were published in the 1970s that address this hypothesis. In the first of these. Gannon and Sternbach (1971) developed a procedure for training alpha activity (measured from the occipital region) in a patient with 3-year history of severe headaches following multiple head traumas. After a little more than 32 hr of training (sixty-seven 29-min sessions), and during no-headache periods, the patient was able to increase his alpha activity from 20% to 92% of the time with eyes closed. He was also able to increase alpha activity to 50%i of the time with eyes open, supporting the efficacy of EEG biofeedback training for making changes in bandwidth activity. However, when he began training during a headache period, lie was unable to concentrate enough to increase alpha activity the headache appeared to interfere with his ability to generate alpha. On the other hand, the intensity and duration of headaches did decrease gradually over the course of treatment for this patient. The patient also reported that after the first 20 sessions, he had a larger attention span and was able to read for 30min without getting a headache (whereas prior to treatment. reading for 15 min induced a headache). Also, following 50 treatment sessions, other activities that prior to treatment induced a headache (swimming, attending concerts) no longer did so.
Andrcychuk and Skriver (1975) treated 33 individuals with migraine headaches with 10 sessions of one of three treatments: hand- warming biofeedback, autogenic relaxation instructions, and 201 alpha enhancement feedback. EEG activity for the alpha (8 13 Hz) enhancement feedback was assessed via bipolar measurement from electrodes placed over the right and left occipital areas, using the right ear as a common ground. Thirty min of training, provided in two 15-min blocks, were provided al each session. Participants in all three treatment conditions, including those in the alpha enhancement group, reported significant reductions in headache rates, and there were no significant differences in improvement between the treatment conditions. Melzack and Perry (1975) recruited 24 patients with a number of different chronic pain conditions (including back pain [// = 10], peripheral nerve injury pain [/> = 4J. and pain from cancer [/i = 3J, among other chronic pain conditions) and provided them with both self-hypnosis and alpha enhancement training (12 participants), hypnosis training alone (6 participants), or alpha enhancement training alone (6 participants). Alpha bandwidth activity and subjective measures of pain were assessed before and after each treatment session. They found larger pre- lo possession decreases in pain (as measured by the MeGill Pain Questionnaire, which assesses different pain qualities and scores them into Sensory and Affective subscales) during training for those participants who received both hypnosis and neurofeedback
Author (Date); Type of Pain
Type of Study
TABLE 1. Summary of findings from neurofeedback studies of chronic pain. N-. No. of Electrode Bandwidth(s) Bandwldth(s) Sessions Placement Reinforced Inhibited
Gannon & Sternbach (1971); Headache following multiple head injuries.
Case report
N— 1 ; 67 29" sessions
Occipital region. right side
Alpha (9-11 Hz)
None
Andreychuk & Skriver (1975); Migraine headache
Treatment comparison (alpha enhancement [*=11]. hand-warming biofeedback [n=11]. autogenic training [n= 11}) Treatment comparison (alpha enhancement alone [n=6]. hypnosis alone [n = 6], both treatments [n=12]).
N= 33; 10 sessions (that included 2 15" training sessions each)
Occipital region, right and left sides
Alpha (8-13 Hz)
None
N = 24; 8 20" sessions
Occipital
Alpha (specific frequen cies not specified)
None
Melzack & Perry (1975); Mixed chronic pain conditions
Findings
Treatment resulted in increases in eyes open alpha, although the patient was un-able to generate alpha during a headache. Intensity and duration of headaches decreased from pre- to post-treatment. - After 20 sessions, the patient reported an increase in attention span, and after 50 sessions, that he was able to attend a rock concert and swim without developing a headache. -Participants receiving alpha enhancement training reported significant reductions in headache rates, as did participants in the other two treatment conditions.
- Participants receiving both neurofeedback and hypnosis reported larger pre- to post-sessions decreases in pain than those that received only one treatment.
Cohen et al. (1980); Migraine headache
Caro & Winter
(2001);
Treatment comparison (forehead cooling/hand warming biofeedback [n=11]. frontalis EMG biofeedback [n=11]. temporal artery vasoconstriction biofeedback (n 10]. alpha enhancement neurofeedback [n-10]). Case series N- 42; 24 20" sessions
N= 15; 40 or more sessions
02 and P4
CZ
Fibromyalgia
Case study Sime (2004); Trigeminal neuralgia N~ 1; 29 sessions of
neurofeedback preceded & followed by EMG/ respe ration
biofeedba ck. Multiple sites used, including T4-A2, C3-A1,
C4-A2. C3-C4. & T3-T4.
Alpha (8-13 Hz)
None
- All participants reported reductions in the number of headaches per week; no between-group differences found. - No changes in headache intensity or duration found. - No changes in alpha activity observed in the neurofeedback group.
SMR
4-7 Hz and
- Participants demonstrated significant
(12-15 Hz)
Variable, including 7.510.5 Hz, 12-15 Hz. 7-10 Hz, 11-14 Hz. 8.511.5 Hz.
22-30 Hz
2-7 Hz and 22-30 Hz
improvements in attention. - There was a strong association between improvement In attention and a decrease (improvement) In the mean tender pain assessment score, and a moderate association between improvement in attention and decrease (improvement) in fatigue. -Training using T3-T4 sites (7.5-10.5 Hz reword. 2 7 and 22-30 Hz inhibit) resulted in the greatest pre- to postsession pain reductions. -The participant reported reductions in pain and improvement in sleep quality -Benefits were maintained at 13-month follow-up. (Continued)
Author (Date); Type of Pain
Type of Study
Af; No. of Sessions
Electrode Placement
TABLE 1 (continued) Bandwidth(s) Reinforced
Bandwidth(s) Inhibited
Findings
Jensen et al. (2007); Complex regional pain syndrome Type 1
Case series
A/-18; analyses focused on pre- to postsession changes.
Multiple sites used. including P3-P4. FP1-FP2, T3-T4. FP02-A2. CZ-FZ. F7-F8.
Variable, ranging from 0-3 Hz to 14-17 Hz.
Not specified
- Significant pre- to postsession improvements were found on patient ratings of pain intensity (at the primary, secondary, and tertiary pain sites), muscle spasm, muscle tension, and global well-being.
Kayran et al. (2007); Fibromyalgia
Case series
N 3; 10 30" sessions
C4
SMR (12-15 Hz)
Theta
- All three participants reported significant preto posttreatment decreases in pain intensity. tatigue, depression, and anxiety. -The participants showed variability in changes in EEG activity; one showed minimal changes and two showed substantial decreases in theta activity
training than in participants who received only one type of training. Moreover, although participants in both of the neurofeedback Scientificconditions Articles showed an increase in alpha output, those who received both hypnosis and neurofeedback demonstrated the most increase in alpha output over the course of treatment, whereas those who received neurofeedback training only demonstrated the most increase in pre- to postsession alpha output. Cohen, McArthur, and Rickles (1980) assigned 42 patients presenting with migraine headache to 24 sessions (over the course of 8-10 weeks) of one of four biofeedback conditions: (a) forehead cooling/hand warming. (b) frontalis EMG reduction, (c) temporal artery vasoconstriction, and (d) alpha enhancement. Alpha EEG activity (8—13 Hz) was measured from 02 and P4. with feedback provided as tone-off (alpha above threshold with the threshold adjusted according to performance) or tone-on (alpha below threshold). All of the participants reported a significant reduction in the number of headaches per week, although there were no changes in the intensity or duration of headaches that did occur, nor were there any significant changes in alpha activity from pre- to posttreatment in participants in the neurofeedback group. Caro and Winter (2001) provided 15 patients with fibromyalgia 40 or more sessions of sensorimotor rhythm (SMR) training (reinforcing 12 151Tz and inhibiting 4 7Hz and 22 30 Hz). The study participants evidenced significant improvement in a visual test of attention (TOVA scores), and there was a strong association between improvements in attention and improvements (decreases) in physician-assessed tender point scores (e.g.. rs — —.64, -.85, and -.69 for the associations between the TOVA ADHD, Commission Errors, and D Prime scores and the tender point score, respectively). Weak to moderate correlations (rs=-.29. -.46. and .16. respectively) were also found between the TOVA scores and patient ratings of fatigue. Sime (2004) presented a case report of a patient with trigeminal neuralgia treated with both neurofeedback (29 sessions) and peripheral biofeedback (10 sessions). The electrode placement and bandwidths reinforced varied as treatment progressed and included training at T4-A2, C3-A1, C4-A2. C3-C4. and T3-T4. The bandwidths reinforced also varied, although 2-7 Hz and 22 301 Iz were consistently inhibited in each session. Sime reported that rewarding low alpha (e.g.. 7.5 10.5 Hz) measured from T3-T4 was associated with the most immediate improvements in pain. Following treatment. the patient chose to avoid a planned surgery (severing the trigeminal nerve) for pain treatment and discontinued the use of an
opioid/acetaminophen combination analgesic. Moreover, the benefits of treatment were maintained in this patient at a 13-month205 follow-up assessment. We recently reported our experience in a retrospective analysis of 18 patients with CRPS-1 who had been given neurofeedback training as part of a muitidisciplinary pain treatment program (Jensen. Gricrson. Tracy-Smith, Bacigalupi, & Othmer, 2007). In the study, participants were administered 0 to 10 numerical rating scale measures of pain intensity at their primary pain site, as well as pain at other sites and other symptoms (e.g., muscle tension), before and after a 30-min neurofeedback training session. The specific neurofeedback training used varied from patient to patient and from session to session, depending on the needs of the patient and goals of each treatment session. Self-reported changes in symptoms were statistically compared using a series of r tests. Across the different treatment protocols, a substantial and statistically significant pre- to postsession decrease in pain intensity at the primary pain site (from an average intensity of 5.2 to 3.2 on a 0 10 scale) was reported, with half of the study participants reporting changes in pain intensity that were clinically meaningful (30% or greater). Five of seven secondary outcome measures also showed statistically significant improvements following neurofeedback treatment. These included pain at secondary and tertiary sites, muscle spasm, muscle tension, and global well-being. Finally. Kayran and colleagues (Kayran. Dursun, Ermutlu. Dursun. & Karamursel. 2007) described a ease series of 3 individuals
206 with fibromyalgia who were given ten 30-min such as periventricular/ peraqueductal gray matter, sessions ofSMR (12-15 Hz activity) training. During training, EEG was assessed from C4, and SMR activity was reinforced and theta activity was inhibited. Each participant reported decreases in pain (absolute pre- to posttreatment intensity rating decreases were 4.0, 1.5. and 3.0 on a 0-10 numerical scale), fatigue, depression, anxiety. However, the 3 participants also showed variability in pre- to posttreatment changes in EEG activity. One participant showed minimal changes in SMR activity, theta activity, or the theta/ SMR ratio. The other 2 participants show no or minimal changes in SMR activity but substantial decreases in theta activity (and therefore associated decreases in the theta/ SMR ratio). As a group, the case and case series reports suggest that neurofeedback training may be associated with decreases in pain and improvement in other symptoms (such as depression and anxiety). Although controlled trials comparing neurofeedback to no treatment (standard care) or placebo treatment have not yet been performed, at least three studies have compared neurofeedback to other established treatments (e.g., hand-warming biofeedback, EMG biofeedback, hypnotic analgesia; Andrcychuk & Skriver, 1975; Cohen et al., 1980*; Melzaek & Perry-, 1975). In each of these studies, the neurofeedback intervention was shown to be at least as as effective as the comparison treatments. When specified, the goal of the neurofeedback training in the published studies has most often been to increase relative alpha activity, although treatment protocols were also sometimes developed to decrease theta and increase in SMR (12-15 Hz) activity. Beta activity was rarely directly targeted, and when it was, the goal was to decrease this activity.
internal capsule, and sensory thalamus, and stimulation of these areas has shown promising results for chronic pain management (Green et al., 2005; Wallace, Ashkan, & Benabid. 2004). Although there have been some promising results using deep brain stimulation, the more common approach is the stimulation of motor cortex. The rationale here is that stimulation of the motor cortex can inhibit pain relays in the thalamus (perhaps by impacting thalamocortical dysrythmia). In fact, investigators have studied the efficacy of electrical stimulation applied directly to the motor strip of the cortex via surgically implanted electrodes and have demonstrated 28% to 47% reductions in pain intensity in patients with chronic pain who have received this procedure (Nguyen et al., 1999; Nuti et al., 2005). However, in addition to the elevated costs, surgical implantation and maintenance of electrodes inside the brain carry substantial risks. Noninvasive brain stimulation techniques can be used instead of invasive ones to address these risks. One such procedure, already mentioned, is rTMS. with which pulses of electromagnetic currents are used to induce electric currents inside the skull. Depending on the stimulation frequency and amplitude, rTMS can stimulate or inhibit activity a focal cortical area. A number of studies have found at least temporary decreases in pain experience in chronic pain sufferers following rTMS applied using inhibitory frequencies to the motor cortex (Lefaucheur. Drouot, Keravel. & Nguyen. 2001; Pleger et al., 2004). Unfortunately. rTMS equipment is expensive and lacks portability, making it less practical than other approaches. In addition, depending on the intensity of stimulation, rTMS can
OTHER TREATMENTS THAT ALTER EEG ACTIVITY ALSO PRODUCE CHANGES IN CHRONIC PAIN Cortical Stimulation If cortical activity is related to the experience of pain, then any intervention that alters cortical activity has the potential to impact the experience of pain. Recent data studying the effects of cortical stimulation for this purpose are promising. There are several techniques to stimulate cortical areas such as invasive (epidural cortical stimulation) and noninvasive approaches using TMS or transcranial direct current stimulation (tDCS). It is also possible to implant electrodes in deep areas JO URNA L OF NEUROTHERA P ) '
be uncomfortable and the procedure is difficult to administer in a blinded fashion. Finally, rTMS induces a strong electric current in the brain that Scientific Articles results in action potentials. It is still unclear whether supra- threshold stimulation is the best approach to modulate conical activity. Another noninvasive technique studied is tDCS. In tDCS. very weak electrical currents (1 to 2 mA) are applied directly onto the scalp via one of two electrodes. Most often, the active electrode is placed over a site of interest, and the other electrode is placed on the contralateral side of the forehead or at an extracephalic area. There is evidence that corneal activity under the scalp where a positive electrode (anode) is placed increases, and activity under a negative electrode (cathode) decreases (Antal. Nitsche, & Paulus. 2001; Nitsche & Paulus, 2001). Additional modeling studies suggest functionally significant amount of electric current may reach the cortex from appropriately large electrodes suitably placed (Miranda. Lomarev. & ITallett. 2006: Wagner et al., 2007). tDCS has promise over the other available stimulation techniques in that (a) it does not require implantation of invasive hardware: (b) it is easy to apply; (c) the tDCS equipment is inexpensive and easy to maintain: (d) given the very low currents involved, active tDCS is very difficult to detect by patients (making a sham-stimulation condition in clinical trials possible); and (e) some evidence suggests that the modulatory effects of tDCS may be stronger than r'lMS (Nitsche & Paulus, 2001). Finally tDCS has an interesting advantage as it modulates spontaneous neuronal firing via modulation of resting membrane potential; therefore, this technique may be suitable to enhance learning effects associated with behavioral tasks. In the context of pain treatment. it is possible to envision the use of tDCS coupled with cognitive restructuring therapy or self-hypnosis training, which might work synergistically to enhance overall treatment effects. Evidence suggests that tDCS holds promise for treating chronic refractory spinal cord injury pain. Fregni and colleagues randomized 17 patients with spinal cord injury and chronic pain to receive sham or active motor cortex tDCS (2 mA, 20min each. 5 consecutive days; Fregni, Boggio, et al.. 2006). They found (a) a significant reduction in pain intensity ratings after active anodal (positive electrode) stimulation of the primary motor cortex in the active, but not the sham, condition; (b) the reductions in pain intensity following each session lasted at least 24 hr until the next treatment session: and (c) there was a cumulative analgesia effect with multiple treatments, such that each treatment produced further reductions in pain from one day to the next.
After 5 days of tDCS, average pain scores decreased more than 50% from baseline (6.2/10 to 2.9/10) in the active group, whereas 207 they actually increased slightly, on average, in the sham group. In another study of 32 patients with fibromyalgia randomly assigned to receive active or sham stimulation over the motor cortex or dorsolateral prefrontal cortex, significant benefits were found only for active tDCS over the motor cortex (Fregni, Gimenes. et al., 2006). This second study supports not only the specificity of tDCS over placebo effects but also the specificity of site placement for the effects of tDCS. However, the extent to which effective tDCS treatment is associated with changes in EEG has not yet been examined. Hypnosis Controlled trials, published over the past decade, have demonstrated that hypnosis training can result in reductions in the severity of both acute and chronic pain (Montgomery. DuHamel, & Redd, 2000; Patterson & Jensen, 2003). Moreover, selfhypnosis training in persons with chronic pain appears to have two primary effects: a short-term reduction in chronic pain that occurs during the treatment session or hypnosis practice that lasts for several hours in about 70% of persons with chronic pain, and a longer term permanent reduction in baseline daily pain, experienced by a smaller subset (about 25%) of patients (Jensen. Barber, et al., 2008). PET and fMRI studies show that the effects of hypnotic analgesia are "real." in
208
JOURNAL OF NEUROTHERAPY the sense that hypnosis with suggestions lor inhibit an individual's ability to be hypnotized, as it analgesia produces reliable reductions in activity in is possible that in some patients, the cortical activity the sensory cortex and other areas of the brain that associated with chronic pain could inhibit some of are associated with the experience of pain and are the necessary conditions for hypnosis. known to process nociceptive information (Hofbauer, Rainville. Duncan. & Bushnell. 2001; Rainville, Duncan. Price, Carrier. & Bushnell. SUMMARY AM) IMPLICATIONS OF RESEARCH ON THE ASSOCIATIONS BETWEEN PAIN AND 1997). However, imaging research has not identified NEUROPHYSIOLOGY a mechanism for the effects of hypnotic analgesia, as it is possible that the observed changes in cortical Research performed over the past decades has activity could be due to a number of factors (e.g.. confirmed that the brain is not a passive recipient of increase control over subsystems that process nociceptive information, but modulates and is itself nociception, distraction). A relatively large number of studies have influenced by nociceptive input. The central examined the EF.G correlates of hypnosis. Two of sensitization that can occur in the CMS with ongoing the consistent findings from this literature are (a) nociception, and that can contribute to ongoing pain, more highly hypnotizable individuals tend to show could also potentially be reversed or interrupted with greater slow wave (theta and alpha) activity than less interventions that alter pain- related cortical hypnotizable individuals, both at baseline (i.e., functioning. This possibility provides a rationale for before hypnotic inductions) and during hypnosis, the use of interventions that target pain modulation and (b) hypnotizable, and especially high at the level of t he cortex, including hypnotizable, persons show an increase in slow nonpharmacological novel interventions that could wave t'theta and alpha) activity following hypnotic potentially benefit patients with chronic refractory inductions (Crawford, 1990; Williams & Gruzelier. pain and with the additional advantage of being a 2001). These findings are consistent with the more specific and focal treatment. Indeed, there has possibility that the presence of slow wave EEG been a dramatic increase in interest in such activity patterns is correlated with the effects of treatments in recent years. Recent, albeit preliminary, research suggests that hypnosis on pain. Also, in one study, neurofeedback the cortical modulation of chronic pain may be training to increase both theta (4-7.5 Hz) and alpha reflected in EEG bandwidth activity, such that (8 12 Hz) activity, as well as the theta/alpha ratio, chronic pain relief is associated with a relative was shown to increase responsivity to hypnotic increase in alpha activity and a relative decrease in suggestions, providing further support for a possible beta. Research also suggests that increased theta impact of hypnosis on EEG-assessed bandwidth activity may be associated with some chronic activity (Batty ct al., 2006). These findings, when neuropathic pain conditions. However, further considered in light of the fact that the majority research (about 70%) of persons who receive hypnosis report decreases in pain, are consistent With the possibility (not yet adequately tested, however) that hypnosis's effects on pain might be directly associated with in EEG-assessed cortical rhythms; specifically because of the relative increases in the slow wave activity that accompanies hypnosis (Williams & Gruzelier, 2001). Of course, even if consistent associations between specific cortical rhythms (e.g.. less beta and/or theta and more alpha) and pain relief arc found, this could not be used as evidence proving that changes in cortical rhythms mediate the impact of some treatments on pain experience. Such evidence must come from experimental studies in which cortical rhythms are systematically altered (perhaps through the use of targeted neurofeedback protocols) in some patients and not others, to determine if changes in certain specific rhythms lead to changes in pain experience. Another important point that needs to be considered is that pain may
is still necessary- to confirm the specific EEG >ignature(s) of chronic pain. Our understandingScientific of the associations between Articles EEG activity and pain is just beginning; much more research is needed to determine which EEG activity patterns, if any, are consistently associated with the experience of chronic pain. However, if future research supports the relationships that preliminary research suggests exist, this research could provide an empirical basis for designing interventions that target brain activity for pain management (such as a neurofeedback) •r be used to guide parameters of stimulation when using tDCS. NEUROFEEDBACK TREATMENT APPROACHES FOR CHRONIC PAIN MANAGEMENT l'his section provides some suggestions regarding how one might utilize neurofeedback in the treatment of chronic pain, based :n large part on the literature review just presented. However, it is also important for the clinician to keep in mind that, because •f the lack of controlled studies, no single protocol or neurofeedback approach has proven efficacy lor pain management at this point in time. In general, an increase in relative rnagni- :udc alpha frequency band (generally defined as 8-12 Hz) is thought to reflect decreases cortical "activation" or directed active engagement. For example, the alpha band has been shown to be associated with an alert vet idle state, or more simply stated, cortical receptivity (Sherlin, 2008). Conversely an increase in relative magnitude beta frequency (generally described as 13-32 Hz) is thought to reflect increased active (dircc- :ed) cortical engagement. We can therefore hypothesize that if the cortical areas associated with the processing of pain can be conditioned to produce decreases in cortical arousal and increases passive receptivity (through decreasing beta and/or increasing dlpha relative magnitude), then the experience of pain may decrease. The setup required for protocols that would increase relative magnitude alpha and decrease relative magnitude beta is easy using most software and hardware systems currently available for neurofeedback application!.. However, questions may be raised concerning the ideal or most practical electrode placement. Two placement protocols that have been described are (a) a T3-T4 sequential montage (i.e., bipolar montage; Sime, 2004) and (2) a C4 placement (Kayran et al.. 2007). Although Sime reported an immediate beneficial effect of training using the T3-T4 montage placement, there is the possibility that this protocol could reduce the
alpha magnitude and increase the relative beta magnitude at the referential site to achieve the training objective (e.g.. 209 T3 active minus T4 reference activity). An additional important knowledge gap in the field that should be recognized is also related to the issue of electrode placement. Alpha frequency activity can be enhanced globally or regionally. Moreover, global and regional changes in bandwidth activity can influence each other and are probably (somehow) related. For example. T3/T4 alpha is likely driven by occipital alpha, which may have something to do with alert relaxation. C3/C4 for frontal alpha, on the other hand, is probably an enhanced Mu rhythm, which can also be achieved by relaxing hand muscles. One can make similar arguments about regional and global influences on beta or gamma bands. More research is needed to determine how comorbidities in the pain patient influence global brain patterns. In the meantime, and until our knowledge concerning these potential mutual influences is increased, or research demonstrates the superiority of one placement over another, it might make sense for clinicians to compare different options (e.g.. T3-T4 vs. C4 vs. other possible sites) in the same patient, and then use the placement that results in the most benefit for that individual patient. Some recent studies have also identified the dorsal anterior cingulate cortcx (dACC: Brodmann Area 24) as critical in the affective experience of pain (e.g., Rainvillect al.. 1997). Recent advanced ncuro-/fMRl-feedback techniques have targeted this location. One study, for example, used fMRl feedback to
JO URNA L OF N EU ROTH ERA P Y demonstrate how the ongoing experience of pain The application of this model for neurofeedback may be modulated by training participants to modify training would be very similar to that used in the dACC activation accessed via real-time fMRI previously cited study addressing the occipital (DcCharms ct al., 2005). Tn another study, an region (Gannon & Sternbach, 1971 ). That is. the advanced cortically targeted neurofeedback peak frequency is first identified in both baseline technique called Standardized Low Resolution conditions of eyes closed and eyes open in the Electromagnetic Tomography (or s LOR ETA) occipital region. This allows the clinician a very targeted this same region with similar results (Ozier. easily detectable and testable hypothesis in Whelton. Mueller, Lampman. & Sherlin. 2008). determining if this protocol might be potentially Although such advanced feedback techniques are useful. In our clinical experience the best electrode not available to most practitioners. one may placement for this purpose is along the midline in the postulate targeting the same areas using parietal and occipital region. Typically we choose conventional neurofeedback with a single channel at the active electrode site of POZ referenced to the ear approximately FZ or just anterior to FZ. or (or other neutral site such as nose) for the eyes open alternatively with two channel referential montage condition; if training in the eyes closed condition we of electrode sites F1 and F2 as active sites. The either use site OZ or a two-channel referential training goal would be to increase the relative montage using sites Ol and 02. The protocol goal magnitude alpha while decreasing/inhibiting relative here again is to increase the alpha frequency relative magnitude beta in the ACC. The clinician should magnitude, with the difference that with this closely monitor client progress of the training, and protocol, we now inhibit both the beta frequency particularly focus the client on being able to identify relative magnitude and the theta frequency relative and replicate any state of comfort achieved during magnitude. training at home, when pain levels are perceived as particularly high. CONCLUSION Although the goal of these protocols is to teach the client to identify and utilize a state of increased Chronic pain is a significant health problem for alpha and decreased beta for the control of pain, many individuals that is not being adequately there is also a potential risk that if the client addressed with the treatments that are currently increases the overall (baseline) alpha and decreased available (Turk, 2002). Moreover, the available beta relative magnitude, he or she could potentially pharmacologic treatments that tend to be used for suffer cognitive deficits (Chabot & Serfontein. chronic pain are often associated with significant 2006). However, we have yet to notice this possible adverse effects (e.g., opioids can lead to tolerance, side effect in the individuals we have treated using constipation, and altered mental states). The this approach for pain management. Nevertheless, understanding that pain experience is modulated at this possibility should be carefully monitored in many levels of the CNS, including the cortex, opens participants in any neurofeedback training protocol. the door to interventions that might affect pain at the As previously noted, in individuals with spinal cortical level, and that may not have as many cord injury and chronic pain, there is evidence for negative side decreased alpha and increased theta frequency band relative magnitude peak, relative to individuals with spinal cord injury who do not have chronic pain. For these patients, a potential goal of treatment would be to shift the peak frequency from the theta frequency band range to the alpha frequency band range to determine if this is associated with improvements in pain. However, the neurofeedback protocol to achieve this goal would be quite similar to that used for increasing alpha and decreasing beta band activity, as both involve increasing alpha. A shift in peak frequency band can be achieved, and has been demonstrated in a variety of populations in our clinical experience, although in our work it has primarily been applied to increase cognitive processing speed and thalamic processing of 210 incoming cortical and sensory information.
effects produced by analgesics. Clinicians knowledgeable in neurofeedback approaches might consider opening their practices to patients with Scientific Articles chronic pain and then reporting to the community the outcomes of their clinical work in the form of published case studies and case series. This sharing of clinical experience provides an important foundation for hypothesis generation, which would then contribute to the design and implementation of more definitive clinical trials. Ultimately, the results of such work would help us to understand the extent to which neurofeedback benefits individuals with chronic pain, and if it does, the specific interventions, protocols. and approaches that are most effective.
Chabot. R. J.. & Serfontein. G. (2006). Quantitative clcctroenccphalographic profiles of children with attention deficit disorder. Biological Psychiatry, 40. 211 951-963. Chang, P. I7.. Arendt-Niclscn. L.. Graven-Nielsen. T., Svenson, P.. & Chen, A. C. N. (2001). Different EEG topographic effects of painful and non-painful intramuscular stimulation in man. Experimental Bran: Research. 141. 195 203. Chen, A. C. N. (1993). Human brain measures of clinical pain: A review. I Topographic mappings Pain, 54, 115-132. Chen, A. C. N. (2001). New perspectives in EEG/ MEG brain mapping and PET/fMRl neuronna- ging of human pain International Journal of Psychophysiology, 42, 147-159. Chen. A. C., Dworkin. S. F., & Drangsholt, M. T. (1983). Cortical power spectral analysis of acute pathophysiological pain. International Journal of Neuroscience, 18, 269-278. REFERENCES Cohen, M. J., McArthur, D. I.., & Rickles, W. H. (1980). Comparison of four biofeedback treatments for migraine Andreychuk, T., 257 Proceedings of the 2009 ISM R Conference Introduction Introduction This study investigated the neurophysiology of object processing in a population of normal university students. We utilized standardized low-resolution electromagnetic tomography (sLORF.TA) to map sources of the EEG recorded at the scalp. /Methods We obtained 100 students (60 female) with a mean age of 21. EEC data were recorded for 4min while participants viewed an image of a hammer. We performed EEG source localization using sLORETA. We compared the image condition to baseline using all voxel by voxel / tests. Significant voxels of difference were mapped onto an MNI atlas containing 6,329. 5 mm voxels. The phenomenology of the recording was obtained at the end of the EEG recordings. Results The contrasts between baseline and object processing show specific differences in regions in the left hemisphere in delta, theta and alpha frequency domains, wheras alpha and beta activity show significant increase in right anterior cingulate (BA 24) and prefrontal regions in addition to right parietotemporal areas. Discussion The left hemisphere appears to play an important role in object processing. The increase is related to the evaluation, categorization. and experiential memories utilized while focusing on the image. The right anterior regions may also play a role in the identification of the object but may also be important to the meaning of the object and its known function in social and intrapersonal contexts. The subjective reports of mental processes and experiences of the participants offer evidence for the patterns of significant increase in these regions. Incremental Gains in Self-Regulation Skills: Comparison After 20 and 40 Sessions of Neurofeedback
EEG biofeedback practitioners around the world have been publishing studies on the effectiveness of neurofeedback training on ADHD. This research has demonstrated that neurofeedback is an efficacious treatment for ADHD, but it has not clarified the number of neurofeedback training sessions that lead to optimal gains. Some practitioners have examined results after 20 sessions of feedback, whereas others have presented results following 40 sessions of training. This pilot project proposes to fill this void by analyzing the effect of the number of sessions of neurofeedback on attention. EEG, and short-term memory (digit span). This study will compare client scores on these measures before neurofeedback training, alter 20 sessions, and after 40 sessions of training. Method Clients range from age 6 to adult, and all met the criteria for diagnosis of ADHD or Asperger's Syndrome according to the Diagnostic and Statistical Manual of Mental Disorders (5th cd.). All clients will have been tested using the Integrated Auditory and Visual Continuous performance test (IVA) at their initial assessment, after 20 sessions of training and after 40 sessions of training. A subset of clients will also have data on EEG changes, W'esehler Intelligence Scale IV digit span subtest data and questionnaire data post-20 sessions and post-40 sessions of
258
JOURNAL OF NEU ROTH ERA F Y
training. There will be a minimum of 20 participants in this study. Training sessions are 50min. and most clients attend twice a week. Training parameters are based on an initial assessment of EEC performed by Dr. Lynda Thompson. Most clients already in the study have been doing training at either Cz or C4 reference to the left ear. The aim for the majority of clients is to decrease slow wave activity (3-7 Hz, 4-8 Hz. or 3-10 Hz), increase sensorimotor rhythm (12-15 Hz or 13-15 Hz) while decreasing any high frequency beta or spindling beta (23-35 Hz). EEG biofeedback training sessions are also combined with me la-cognitive strategies.
Results This study is not yet completed. Results will include the 1VA. questionnaire data (Conners' Global Index for ADHD. ADD Centre's ADDQ. and DSM1V questionnaire for ADHD), and EEG. EEG ratios include thcta/beta and theta/SMR ratios.
Net corticospinal excitability of the left hemisphere was significantly increased more than 20 min after the end of alpha suppression (desynchronization), as reflected in the average magnitude of the motor evoked potential (MEP; 130% of baseline), together with a significant reduction of short intracortical inhibition (165% of baseline). Of importance. MEP change was inversely correlated with percentage of alpha amplitude change during NFB (r> 0.5, p < . 05), as well as with the ratio of pre-to-post alpha baseline at rest. Following low beta NFB training there was a significant enhancement, of intracortical facilitation, without reliable main effects in MEP amplitude, both seemingly a result of uneven entrainment. Nevertheless a significant negative correlation was observed between the magnitude of low beia synchronisation and single-pulse MFP change (r>—0.5, /; cation of specific computerized attention training exercises in conjunction with neurofeedback has Introduction been shown to significantly improve attentional functioning for both mTBI and adult ADHD Neurofeedback training integrated with other populations in only 20 training sessions. In addition, behavioral techniques could be one of the potentially cognitive training prov ides an ethically appropriate efficacious intervention options for cocaine control treatment for neurofeedback research. These addiction treatment. Our study combined SMR studies and additional research incorporating neurofeedback treatment with motivation meta-cognition training in identifying effective enhancement therapy for the treatment of outpatients means for the integration of cognitive rehabilitation with cocaine addiction. EEG changes in beta and and neurofeedback are reviewed in this paper. theta power are typical for withdrawal from cocaine. Executive prefrontal functional deficits have been for both active users and recovering addicts. Brain Oscillatory Correlates of Visual Attention and reported We proposed that cocaine abusers may benetlt from Short-Term Memory SMR and SMR/Theta neurofeedback protocol. Paid Sauseng. PhD University Motivational interviewing techniques were of Salzburg employed to engage outpatient subjects in neurofeedback and retain them during 12As we are continuously bombarded with visual session-long neurofeedback training information, we have to select relevant from Cognitive test based on Eriksen Hanker taskcourse. with irrelevant visual input. Based on a recent hypothesis dense-array event-related potential (ERP) recording on the functional meaning of used to assess intervention effects on such electroencephalographic (BEG) alpha activity, local was executive functions as cortical inhibition, motor 10-Hz oscillations are discussed as a correlate of response conflict and error monitoring efficient suppression of irrelevant visual information. along with moredetection, traditional clinical outcome Mortens, R., & Allen. J. J. B. (2008). lhe role of psychophysiology 111 forensic assessments: Deception detection, ERPs. and virtual reality mock crime scenarios. Psychophysiology, 45. 286-298. Rosen fold,.). P., Angell. A., Johnson, M., & Qian, J. (1991). An PRP-based. control-question lie dclcctor analog: Algorithms for discriminating effects within individuals' average waveforms. I'sychopHysiology, 38, 319-335. Roscnfcld, J. P., Cantwell. G-, Nasman, V. T.. Wojdac, V., Ivanov, S.. & Mazzcri, L. (1988). A modified, event-related potential-based guilty knowledge test. International Journal of Neurosdence, 24, 15? 161
Proceedings of lite 2009 ISNR Conference
261
measures. We report immediate post-treatment maintained, whereas the ERN measure did not show effects in 14 subjects and 6 months follow-up effects significant difference from the intake level. Among in 11 subjects. the clinical outcome measures the most significant was a decrease of depression scores (BDI II, p = .01) and a decrease of PTSD scores (p — .02). Depression Method scores on follow-up remained lower than at the level (pre. 25.8; post, 14.7; follow-up, ERPs were acquired with a 128 channel Electrical pretreatment 12.5, follow-up vs. pretreatment, p < . 05). The drug Geodesies Inc Net Station EEG device prior and screens did not showsignificant decrease in cocaine following 4 week long bio-beliavioral intervention use posttreatment; however, of positive tests using a speeded forced-choice reaction time task for marijuana use decreasednumber siguificantly (post- vs, (Eriksen danker tests) with NoGo trials. Follow-up pre-NFB urine drug screens: cocaine test was conducted after 6 months. Beside behavioral nonsignificantly decreased by 14%. p = use. A6. and ERP measures during flanker test, the treatment marijuana use rate decreased by 71%, ¿>
